Social media: “Children are particularly at risk”

Should there be limits on how much children are allowed to use social media? Or is it more a matter of providing them with better guidance? Researchers from various disciplines at LMU explain how Instagram and other platforms capture their users’ attention and what that means for children and adolescents.

Why do social media platforms encourage people to spend so much time on them? What does this mean for children and adolescents?

Enlarge

“Social media channels are a nice idea in principle – they make it easy to connect and communicate with other people around the world. There’s always something new to discover. But studies also show that users regularly spend more time on social media than they really intended to. And often they don’t feel better but actually feel worse after using social media. They experience feelings of envy or anger, or end up finding the content fairly trivial.

“There are many different reasons why people keep coming back to social media and find it hard to stop. Both the design of the platform and human psychology play a role. First, there is no natural endpoint. Endless feeds encourage users to keep scrolling endlessly, while autoplay automatically starts the next video. The constant visibility of other people’s activity intensifies the feeling that you might be missing out. A person’s dopamine system will go into overdrive, creating the expectation that something interesting is just about to appear – something worth hanging around for.

Resisting all these mechanisms is a major challenge for anyone – and it’s even harder for children and adolescents.”

Sarah Diefenbach is Professor of Economic Psychology and Human-Computer Interaction at LMU.

How safe is it for children to use social media and AI?

Enlarge

“Compared to adults, children are exposed to a particular level of risk on social media. The dangers range from feelings of inadequacy and bullying to grooming and extortion. Artificial intelligence is expanding the capabilities of malicious actors online and on social media, with hate campaigns, highly targeted manipulation, and deepfakes in text, images, audio, and video. As is the case with online content in general, people should be cautious about uploading any private data to AI tools – depending on the provider and its policy, this data may be used for training purposes. Employees may also have access to the data, as is the case with most cloud services.

Children need to learn to critically scrutinize social media content and AI-generated output – even more than before. Chatbots like ChatGPT can sound very convincing, even when their output is wrong. Another fairly new phenomenon is people forming an emotional attachment to chatbots – this affects adults as well as children.

In general, parents especially need to pay careful attention to what their children are doing online and with AI tools. A regulatory measure that could be implemented in the short term would undoubtedly be a ban on children using social media. I would go even further and say that, as a basic rule, children and adolescents under the age of 16 should not have unrestricted access to the open Internet. They can also be taught how to use media and AI tools through specially designed, age-appropriate programs. This would of course all be difficult to enforce, but a ban would also send a signal to parents.”

Johannes Kinder is professor and Chair of Programming Languages and Artificial Intelligence at LMU.

Do parents need to be more involved in supervising their children’s use of AI and social media?

Enlarge

“One problem with the current debate about banning social media for children and adolescents is the highly one-sided assumption that limiting access alone will protect them from the challenges of navigating an increasingly digital world. Rather a holistic perspective is needed: one, that considers not only possible restrictions, but also measures on how to guide and empower young users.

Previous research on parental media education shows that a restrictive approach, such as limiting content and screen time, can minimize exposure to online risks. However, only active media education from parents – for example, open conversations and time spent reflecting together – will also allow children and adolescents to benefit from the opportunities that digital media can offer.

In addition to parents, educational institutions also have a role to play here: they can reach a wide range of children and adolescents and teach them how to be media-literate even though they may all be growing up in very different media environments. Rather than focusing primarily on age limits, the debate should focus more on individual risks and protective factors that enable children and adolescents to develop a healthy relationship with digital media.”

Professorin Ruth Wendt is Professor of Communication Science with a focus on “Digital Literacy in Algorithmic Spaces” at LMU’s Department of Media and Communication.

Enlarge

Children need not only protection, but also sound judgment

“New technologies change specific behaviors. But they also change fundamental ways of living together. Cell phones and smartphones, for example, have shaped how we communicate and make ourselves available to one another. It makes a difference whether someone cannot be reached for several days or whether they can, in principle, be contacted at very short notice, and whether an immediate emotional response can be expected. Developments in AI will also bring about changes of this kind. One point is important, however: Children are not merely passive recipients of these technologies. They adopt them in their own ways. Experience with earlier technologies shows that they sometimes use them differently from what adults or providers intended.

There is therefore no generally valid answer to what using AI means for children or how individual children engage with it. AI and social media already encompass too many different aspects for that to be the case. A more precise assessment must take into account factors such as the child’s age and stage of development, the type of AI application and the context in which it is used.

Precisely because of these diverse criteria and the rapid pace of technological change, children’s ability to reflect and make increasingly autonomous judgments needs to be strengthened: What is good for me in the long run? Who can I trust? What consequences do my actions have for others? These abilities, however, do not develop in isolation. They emerge through patient dialogue and shared reflection within social relationships, such as at school and in the family. In these settings, children can learn how to use different technologies and services and be encouraged to reflect on specific experiences.

Families and schools should not be left to manage these issues alone. As a society, we should hold providers accountable for designing their systems in ways that do not undermine the development of independent judgment—for example, by seeking to create emotional dependency or keep users engaged for as long as possible.”

Markus Paulus is Chair of Developmental Psychology and Educational Psychology at LMU.

Enlarge

Should children's use of social media and AI be restricted?

“During adolescence, the reward system is particularly sensitive to rewards and stimulation, while the ability to control impulses and regulate attention is still maturing into the early twenties. And since a phone can reduce attention simply by being present, a thirteen-year-old with a smartphone in their pocket during class is already paying an attentional cost, even if they never unlock it.

So this is not just a “use it responsibly” problem. We already restrict alcohol, gambling, and driving by age for similar reasons. Given what we know about attention and the developing brain, it seems reasonable to consider similar protections in the case of social media.

With generative AI, we face a related concern: these tools can take over the cognitive work (reasoning, writing, problem-solving) that young brains need to practice in order to develop. Outsourcing that effort may seem efficient, but during adolescence, it may come at a real developmental cost. I am not advocating for restricting AI from education; rather, the question is how we design AI tools that support learning instead of replacing it. An AI that quizzes you, challenges your reasoning, or asks you to explain your thinking is fundamentally different from one that simply hands you the answer. “

Professor Zhuanghua Shi leads the Multisensory Perception Lab at the Chair of Neuro-Cognitive Psychology at LMU

Who is criminally liable when children misuse AI or engage in criminal conduct on social media? Conversely, how can children be protected?

Enlarge

“When following the current debate about the risks associated with children’s use of AI and social media, lawyers may occasionally feel as though they are living in a parallel universe. After all, the legal framework is quite clear: Young people aged 14 and older who feel the need to use information technology to insult, coerce, threaten, bully or stalk others — typically other children and adolescents — must answer for their actions under criminal law just as adults do. The same applies to inciting others to commit crimes, disseminating right-wing extremist ideology or sharing material that incites hatred.
Until they reach the age of majority, however, they remain subject to juvenile criminal law, which is guided by the concept of rehabilitative education. Before using the proverbial ‘sledgehammers’ of adult criminal law — fines and imprisonment — to crack adolescent ‘nuts’, the justice system therefore first tries educational measures. There are few limits to the creativity of socially inclined judges in this area. Some judges, for example, summon young offenders who have distributed swastikas to the local court and watch all three hours and seventeen minutes of Schindler’s List with them. While that will certainly not have an effect on everyone, it does prompt many to reflect on their actions.

Parents and other legal guardians also bear responsibility. The familiar warning often seen around construction sites — ‘Parents are responsible for their children’ — also applies in criminal law. Lawyers refer to this as parents’ guarantor status: Anyone who stands by and watches their children commit crimes over the internet, despite being in a position to prevent them from doing so, is in effect regularly committing a crime of omission themselves. In legal terms, children who commit crimes are therefore ‘sources of danger’ that their parents are required to keep under control.

Above all, however, parents — supported by schools — have a role to play in protecting children from the risks of AI and social media. There is no getting around having the relevant conversation, even though it is generally no one’s idea of fun. Someone has to teach young people that the supposedly amusing meme featuring the mustachioed ‘Mr. Adolf’ or jokes targeting people because of their skin color are not funny at all but pose a serious threat to our society.

This requires a great deal of time and effort. Anyone with teenage children knows how much work it takes to draw them out of the filter bubbles on their smartphones and offer them compelling alternatives in the offline world. The ‘big tech’ companies have managed to persuade us that without social media, it is no longer possible to communicate, join the conversation or participate in society. This is clearly nonsense. Books, telephones, e-mail, pens, stationery, sports fields and playgrounds still exist. And the world beyond the virtual realm still has plenty of interesting things to offer. Perhaps we should simply begin by spreading this straightforward message before immediately resorting to bans and penalties.”

Professor Mark A. Zöller holds the Chair of German, European and International Criminal Law and Criminal Procedure, Business Criminal Law and Legal Tech.